Neonatal Renal Vein Thrombosis: Review of the English-Language Literature Between 1992 and 2006
Renal vein thrombosis is a complication that occurs in neonates with various underlying risk factors. It carries a grave prognosis for affected kidneys. Anticoagulant and fibrinolytic therapies have been promoted in the past with anecdotal success in some circumstances. However, prospective controlled trials are still lacking, and to date there have been no evidence-based guidelines available for the treatment of neonates with renal vein thrombosis. We retrospectively reviewed all the available medical literature pertaining to renal vein thrombosis published in English during the past 15 years. A total of 271 patients from 13 case series were identified by using the terms “renal vein thrombosis” and “neonates” via PubMed and Cochrane Library searches. Data then were extracted from each of the studies for analysis. During the past 15 years, a male predominance (67.2%) in neonatal renal vein thrombosis has been reported. More than 70% of patients had unilateral renal vein thrombosis, which was more prevalent on the left side (63.6%). The thrombus involved the inferior vena cava and was associated with adrenal hemorrhage in 43.7% and 14.8% of neonates, respectively. Forty percent of the patients were treated conservatively with supportive care alone. Among those patients who received anticoagulation therapy, unfractionated heparin and low molecular weight heparin were used alone in 21.6% and 20.7% of the patients, respectively. Fibrinolytic treatment alone was used in 11.2% of the patients. Only a minority of patients were treated with antithrombin (1.7%), warfarin alone, (0.9%) or underwent surgical intervention (0.3%). The majority (70.6%) of the involved kidneys became atrophic. A total of 9 neonates died with non–renal vein thrombosis–related conditions during the study period. Evidence-based recommendations on treatment cannot be made at the present time. Cooperative prospective studies that involve multiple centers are needed to elucidate the optimal treatment for neonatal renal vein thrombosis.
The exact incidence of renal vein thrombosis (RVT) in neonates is difficult to determine, because large-scale population-based epidemiologic studies are lacking. It has been reported to occur in 0.5 per 1000 admissions to NICUs according to a large international registry.1 On the other hand, a recent study from Germany suggested that the incidence of symptomatic RVT in neonates was at least 2.2 per 100000 live births.2 Since the systematic and detailed review on the hemostatic complications of pediatric renal disease by Andrew and Brooker3 more than 10 years ago, multiple case series and reports on neonates with RVT have been published. We reviewed the available English-language medical literature pertaining to neonatal RVT over the past 15 years to analyze and examine the data that are currently available on this neonatal disease. We intended to evaluate the efficacy of our current management strategies and identify potential improvement in the care of neonates with RVT.
We searched the PubMed database from the National Library of Medicine using the keywords “renal vein thrombosis” and “neonates” with the limits set to only English-language articles and those that involved human subjects. An additional search was performed via the Cochrane Central Register of Controlled Trials and the Cochrane database of systemic reviews (Issue 4, 2006). We included all case reports and case series that were published in English from January 1992 to December 2006 that reported neonates with RVT. We excluded studies that contained discussion related to neonatal RVT but did not report on an actual case and those reports with 2 or fewer patients. Because the diagnostic criteria of RVT were not clearly stated in every study, we could not scrutinize the accuracy of the diagnosis of each reported case.
Data were extracted from individual studies onto a spreadsheet (Excel; Microsoft, Redmond, WA) for additional analysis. Because the design and methodologies used varied among the studies, some did not have all the information that we intended to analyze. Data are included in the analysis if they were actually described in the study; if a test was not mentioned in the study, the result was entered as unknown.
The search strategies identified 77 publications; 64 were excluded from analysis. Thirty-seven of those reports were excluded because the number of cases included were 2 or fewer.4–40 One case series, reported by the 1-800-NO-CLOTS registry,41 was also excluded from the study because of the potential overlap of patients from other reported cases. Another case series of 23 children with RVT was also excluded42 because 4 of the patients presented after the neonatal period but could not be removed from the aggregate analysis. As a result, a total of 13 case series were included for analysis.1,2,43–53 All of the reported cases had objective documentation of RVT from ultrasound.
Table 1 describes the studies that were included in our review. A total of 271 neonates were included, with a male predominance (67.2% [127 of 189]). Figure 1 depicts the time of onset of RVT. Although all the patients were diagnosed to have RVT within the first month of life, detailed information on the exact time of disease onset was not available from all studies. However, data compiled from studies with that information show that 7.3% (6 of 82), 67.1% (55 of 82), and 25.6% (21 of 82) of the neonates presented in utero, within 3 days, and more than 3 days after birth, respectively. Most of the patients were born at term (71.3% [97 of 136]).
Of the neonates, 70.3% (173 of 246) had unilateral RVT; 63.6% (63 of 99) of these cases involved the left kidney. Figure 2 depicts the features of the patients at presentation. The thrombus extended into the inferior vena cava in 43.7% (90 of 206) of the patients. Adrenal hemorrhage was present in 14.8% (31 of 210) of the patients. Because most of the patients were reported as case series rather than as individual cases, individual clinical features at presentation could not be analyzed and correlated with outcomes. Perinatal risk factors including asphyxia were identified in 31.9% (69 of 216) of the reported cases. Other known risk factors such as maternal diabetes mellitus and dehydration were reported in 8.1% (17 of 210) and 1.5% (3 of 206) of the patients, respectively. Although all the patients had ultrasound-documented RVT, most of them also had at least 1 of the 3 cardinal signs of RVT at presentation: 56.2% (73 of 130) had macroscopic hematuria, 45.4% (83 of 183) had a palpable abdominal mass, and 47.5% (85 of 179) had thrombocytopenia.
Prothrombotic Risk Factors
Not all of the studies collected data on their patients' prothrombotic risks. Before 1996, the only prothrombotic abnormality reported was the presence of the lupus anticoagulant. Since that time, other prothrombotic factors, including protein C, protein S, plasma antithrombin III activity, lipoprotein(a), factor V Leiden mutation, prothrombin gene mutation, and methylenetetrahydrofolate (MTHFR) thermolabile mutation have also been screened in some studies. Among patients with RVT in whom prothrombotic factors were investigated, 53% (79 of 149) had at least 1 risk factor identified. The recurrence rate of thromboembolic events was reported in 1 study.47 With a median follow-up time of 4 years (range: 0.6–15 years), 6.8% (4 of 59) of the patients with RVT had a second thrombotic event, occurring during puberty in 3 of the 4 patients. All 4 patients had at least 1 prothrombotic risk factor.
Ten of the 13 studies reviewed contained data on treatment of RVT. Figure 3 illustrates the different treatment modalities that the patients received: 39.7% (92 of 232) received supportive therapy alone, and 21.6% (50 of 232) and 20.7% (48 of 232) of the patients received unfractionated heparin or low molecular weight heparin (LMWH) alone, respectively. Slightly more than 11% (11.2% [26 of 232]) were given fibrinolytic treatment. A small minority of patients were managed with antithrombin (1.7% [4 of 232]) or warfarin (0.9% [2 of 232]) alone or surgical interventions (0.3% [9 of 232]). Approximately 4% (3.9% [9 of 232]) of the patients received combinations of treatment that included heparin, fibrinolytic agents, warfarin, and protein C concentrate.
Renal outcomes were not reported consistently in the studies. Regardless of the treatment received, irreversible damage was found in 70.6% (156 of 221) of the affected kidneys at last follow-up. In patients treated with unfractionated heparin/LMWH or supportive care, 75.3% (61 of 81) and 72.5% (45 of 62), respectively, of the affected kidneys were found to be atrophic at last follow-up (Table 2). Although the definitions used for chronic renal insufficiency were heterogeneous and often unspecified, 8 neonates (3.0% [8 of 271]) required chronic renal replacement therapies and renal transplantations. Four of these 8 neonates had bilateral RVT; the extent was not documented for the other 4 patients.
Seven of the 13 studies provided follow-up information for hypertension: 19.3% (27 of 140) of the patients had persistent elevation of blood pressure. Laterality of the involved kidneys was provided for 120 patients: 18.9% (14 of 74) of the neonates with unilateral RVT and 21.7% (10 of 46) of the neonates with bilateral RVT had persistent elevated blood pressure (Table 3).
A total of 9 deaths were reported, all of which were related to coexisting medical conditions such as respiratory failure in 3 neonates, multiorgan failure in 1 neonate, and sepsis/meningitis in another neonate. The cause of death of the other neonates was not provided by the authors of the report.
RVT is the most common non–catheter-related thrombosis in infancy and occurs primarily in the newborn period.3 Although many reviews of RVT can be identified in the literature, evidence-based management guidelines are still lacking. In this study, we retrospectively analyzed all the available case series with more than 2 patients on neonatal RVT published in English over the last 15 years. Our intent was to gather the accumulated experience from different sources, evaluate the current management strategies for neonatal RVT, and provide direction for future research in this important hemostatic disorder in neonates.
In an earlier review that included 268 children with RVT, of which 212 were neonates, the incidence in boys and girls was similar and the left and right sides were affected equally.3 Other studies have noted a male predominance, as we have observed in this review. It is uncertain whether this represents a changing demographic of RVT. In our study, the incidence of bilateral RVT was 29.7%, compared with 24% in the earlier review. Preterm infants have been considered to be at risk of developing RVT, and we found that 28.7% of the affected neonates were born before 36 weeks of gestation. Although the classical “triad” of RVT (macroscopic hematuria, palpable abdominal mass, and thrombocytopenia) has been well described, these 3 elements are not always found at presentation. Winyard et al44 recently showed that only 23.5% (5 of 21) of their neonates had the triad at presentation. Perinatal asphyxia, dehydration, and maternal diabetes mellitus are established risk factors for RVT. However, we found that less than one third of the affected neonates had a history of perinatal asphyxia, and dehydration and maternal diabetes mellitus were even less common.
Hereditary prothrombotic risk factors may also play a role in the pathogenesis of neonatal RVT, and routine screening for known procoagulant abnormalities has been suggested.24,45 We found that 53% (79 of 149) of the patients who had been investigated were found to have 1 or more prothrombotic risk factors. In a recent study of 301 children with various types of spontaneous thromboembolism, 21.3% of the patients experienced a recurrence at a median of 3.5 years after cessation of prophylaxis.54 Among those children whose symptoms recurred, 48.4% had a single prothrombotic risk factor and 46.9% had 2 or more risk factors. The authors concluded, therefore, that it is reasonable to screen children with symptomatic thromboembolic diseases for prothrombotic risk factors. However, 1 study in our review47 found that only 6.8% (4 of 59) of the affected neonates had a recurrence of thrombosis, and 3 of the 4 patients had the recurrence during puberty. Another study did not find any recurrence of the thrombosis beyond the neonatal period in neonates with prothrombotic risk factor at a median follow-up of 3.7 years (range: 0.5–20.2 years).45 Because there are no data to indicate that routine screening in neonates with RVT reduces the recurrence risk, there is no evidence to support routine screening for prothrombotic risk factors in neonates with RVT.
Ultrasound is a useful and convenient clinical tool for the diagnosis of RVT. Sonographic features of RVT include enlarged and echogenic kidneys with attenuation or loss of corticomedullary differentiation. Calcification and thrombus may be seen extending outside the kidneys to the inferior vena cava. Ultrasound may also be of prognostic significance for RVT. Winyard et al44 reported a negative correlation between renal length and renal outcomes. Doppler studies are particularly useful for detecting the resistance or absence of flow in renal venous branches and collateral vessels. Although the blood flow in the main renal vein and its branches may be normal, there may be an increase in resistance in the renal arteries caused by thrombosis in the small intrarenal veins. Renal scarring and atrophy are well-recognized features after RVT in affected kidneys, which can also be assessed by a radionuclide scan.
Heparin therapy has been proposed for neonates with RVT, especially in those with bilateral RVT and with inferior vena cava involvement.3 Recently, the trend has been toward the use of LMWH rather than unfractionated heparin. However, almost 40% of the affected neonates were managed by supportive care alone. Our data show that the treatment of neonates with RVT varies greatly among centers, and evidence-based recommendations on this practice are not available. The data suggest that the renal outcomes were similar between supportive treatment and heparin therapies, and a similar proportion of affected kidneys became atrophic in neonates who were managed supportively or with heparin. RVT in neonates often leads to irreversible damage, and anticoagulant therapies may not have an impact on the long-term outcomes.
Although there is no consensus on the management of neonatal RVT, the management should involve a multidisciplinary team that includes neonatologists, radiologists, hematologists, and nephrologists. During the acute phase, supportive management of electrolytes and fluid balance with input from the nephrologists is essential, particularly if the neonates are in acute renal failure and require renal replacement therapies. The hematologists should be consulted on whether and when the neonates require anticoagulation or fibrinolytic treatments. The response to the treatment should be monitored by the clinical team in conjunction with the radiologists and ultrasonographers.
Affected neonates must be followed closely for renal complications such as hypertension, atrophy, functional loss, and chronic renal insufficiency. In our review, reporting on outcomes of the affected neonates was incomplete in many of the publications; the length of follow-up and definitions for chronic kidney diseases were also variable. Persistent hypertension was reported in one fifth of the neonates, and 8 neonates (3.0%) were reported to need chronic dialysis or kidney transplantation. Because of the heterogeneity of treatment approaches and outcome measures, we are not able to draw any conclusion on the impact of different treatment strategies on long-term outcomes.
Nine deaths were reported, and all were non–RVT related. The mortality rate in children with thromboembolic disease from all causes was reported as 5% in a large international registry1 that was published more than 10 years ago. The mortality rate of neonates with RVT that we observed in our study was 3.3% (9 of 271).
Although we retrospectively pooled the available data on neonates with RVT over the last 15 years from multiple centers, we still cannot make any evidence-based recommendations for current treatment strategies. Randomized, controlled trials may not be feasible because of the low incidence of RVT and limited patient numbers in individual centers. However, a prospective multicenter trial with standardized protocols for treatment and long-term follow-up of neonatal RVT should be possible and practical.
Our study illustrates the variability of treatment strategies for neonates with RVT and, potentially, a change in demographics of this condition. Renal outcomes of the affected kidneys are still unsatisfactory for the majority of neonates with RVT. Evidence-based recommendations for the optimal treatment of neonatal RVT are not possible, but prospective multicenter trials with standardized protocols are feasible and should be undertaken as soon as possible to identify the most appropriate treatment strategies for neonates with RVT.
- Accepted April 24, 2007.
- Address correspondence to Anthony K.C. Chan, MD, McMaster University, Department of Pediatrics, Division of Hematology/ Oncology, Room 3N27, 1200 Main St W, Hamilton, Ontario, Canada L8N 3Z5. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Schmidt B, Andrew M. Neonatal thrombosis: report of a prospective Canadian and international registry. Pediatrics.1995;96 :939– 943
- Shafagh H, Soltani MA, Douvoyiannis M, et al. Index of suspicion. Pediatr Rev.2005;26 :461– 466
- Klinge J, Scharf J, Rupprecht T, Boswald M, Hofbeck M. Selective thrombolysis in a newborn with bilateral renal venous and cerebral thrombosis and heterozygous APC resistance. Nephrol Dial Transplant.1998;13 :3205– 3207
- Formstone CJ, Hallam PJ, Tuddenham EG, et al. Severe perinatal thrombosis in double and triple heterozygous offspring of a family segregating two independent protein S mutations and a protein C mutation. Blood.1996;87 :3731– 3737
- ↵Veiga PA, Springate JE, Brody AS, Cummings JJ, Mosovich L, Feld LG. Coexistence of renal vein thrombosis and adrenal hemorrhage in two newborns. Clin Pediatr (Phila).1992;31 :174– 176
- ↵Messinger Y, Sheaffer JW, Mrozek J, Smith CM, Sinaiko AR. Renal outcome of neonatal renal venous thrombosis: review of 28 patients and effectiveness of fibrinolytics and heparin in 10 patients. Pediatrics.2006;118(5) . Available at: www.pediatrics.org/cgi/content/full/118/5/e1478
- ↵Winyard PJ, Bharucha T, De Bruyn R, et al. Perinatal renal venous thrombosis: presenting renal length predicts outcome. Arch Dis Child Fetal Neonatal Ed.2006;91 :F273– F278
- ↵Kosch A, Kuwertz-Broking E, Heller C, Kurnik K, Schobess R, Nowak-Göttl U. Renal venous thrombosis in neonates: prothrombotic risk factors and long-term follow-up. Blood.2004;104 :1356– 1360
- ↵Nowak-Göttl U, Junker R, Kreuz W, et al. Risk of recurrent venous thrombosis in children with combined prothrombotic risk factors. Blood.2001;97 :858– 862
- Copyright © 2007 by the American Academy of Pediatrics